Coordinate systems do not match

[anawas]

Summary

The ALMA observatory (Atacama Large Millimeter/submillimeter Array) can work in different configurations. For most of them there are OSKAR .tm-files available in Karabo. If one plots the configuration of, say version 'cycle 8.10' and 'out 09' with karabo the coordinates of the stations differ significantly (see picture). I reckon that the coordinates in image 'ALMA Version Cycle 8.10' are correct. But the ones in the image on the right hand side are not. The centre is much too far away from the stations.

Steps to reproduce

Run the follwing script and compare the output images:

from karabo.simulation.telescope import Telescope, ALMAVersions

site_name = "ALMA"
instrument_versions = [ALMAVersions.CYCLE_8_10, ALMAVersions.OUT_09]
for version in instrument_versions:
    alma_site = Telescope.constructor(site_name, version=version)
    alma_site.plot_telescope(f"{site_name}.{version.value}.png")

Expected result

The coordinate range of both images should be alike.

[anawas]

Possible cause

I looked at the .tm-folders in karabo/data. It seems that the layout data has different units.
For alma.cycle8.10.tm/layout.txt one finds:

-1.137979806999999937e+03 9.995963548000000642e+00
3.863476964999999836e+02 -1.962095364999999902e+03
4.482469952999999805e+02 6.461269628000000012e+02
-1.275851495000000114e+03 -1.433665864999999940e+03
1.241720548999999892e+03 -1.079905150999999933e+03
...

I assume this is in metres from centre. For alma.out09.tm/layout.txt there is

6.277898100000000559e+05 7.453079620000000112e+06
6.278013100000000559e+05 7.453100269999999553e+06
6.278142600000000093e+05 7.453128290000000037e+06
6.278603100000000559e+05 7.453071139999999665e+06
6.278348499999999767e+05 7.453132370000000112e+06
...

This looks like geocentric coordinates.

[mpluess]

Nice find. Here are all details about the OSKAR telescope model format, might be helpful: https://ska-telescope.gitlab.io/sim/oskar/telescope_model/telescope_model.html

[anawas]

The coordinates seem not to be geocentric ones. Converting them to geodesics point to locations that are not in the vincinity of the observatory.
The files in question were added 2 years ago. Unfortunately, I cannot find a reference to where the data comes from. Does anybody have an idea where they were found?